Pith. sign in

REVIEW 4 major objections 6 minor 45 references

Holistic Artificial Intelligence in Medicine; improved performance and explainability

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A generative-AI preprocessing layer raises average clinical AUC from 79.9% to 90.3% in the xHAIM framework.

desk verdict The 10-point AUC gain is likely an artifact of a baseline that was copied from the original HAIM paper rather than re-run; the framework idea is still worth refereeing. read the letter →

arxiv 2507.00205 v1 pith:RAHIISBW submitted 2025-06-30 cs.AI cs.LG

Periklis Petridis , Georgios Margaritis , Vasiliki Stoumpou , Dimitris Bertsimas This is my paper
classification cs.AIcs.LG
keywords explainableAImultimodalclinicalpredictiongenerativepatientsummarizationHAIMdecisionsupportretrieval-augmentedLLMevaluation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to establish that a generative-AI preprocessing layer can make an existing multimodal clinical prediction framework both more accurate and more explainable. Its proposed system, xHAIM, first selects patient-record chunks most relevant to a given task, summarizes them with an open-source LLM, then feeds the summaries through a fine-tuned ClinicalBERT into the HAIM predictor. On the public HAIM-MIMIC-MM dataset, the authors report that average AUC rises from 79.9% to 90.3% across chest pathology detection, 48-hour mortality, and length-of-stay prediction, with the largest gains in pathologies whose detection depends on clinical narrative. They also report that the generated explanations, which cite specific patient documents, score comparably under human and LLM-based evaluation. If correct, this would mean that filtering and summarizing clinical data before prediction is more effective than feeding models unfiltered multimodal records.

What carries the argument

The load-bearing object is the four-step xHAIM pipeline: (1) retrieval of task-relevant chunks via a hybrid score $\text{Score}_{\text{hybrid}} = \alpha\cdot\text{BM25}_{\text{normalized}} + (1-\alpha)\cdot\text{SBERT}_{\text{sim}}$ with $\alpha=0.5$; (2) LLM-generated task-specific summaries of the selected chunks, using Llama-3.3-70B or Qwen3-32B, with chest X-ray images converted to text descriptions by Qwen2.5-VL; (3) per-modality fine-tuned ClinicalBERT embeddings concatenated with tabular and time-series features into $X = [x_{\text{notes summary}}, x_{\text{cxr summary}}, x_{\text{tabular}}, x_{\text{time series}}]$ and fed to XGBoost; and (4) explanation generation that grounds each prediction in patient summaries, calibrated outputs, and medical knowledge with document citations. The key mechanism claimed is that filtering and summarizing removes noise from long clinical notes, allowing fine-tuning to concentrate on signal.

What would settle it

Re-run the original HAIM pipeline on the same five 80/20 splits used for xHAIM; if its average AUC matches the published 79.9% rather than xHAIM's splits, the reported comparison is not controlled. Separately, feed the fine-tuned ClinicalBERT and Qwen2.5-VL features the full unfiltered notes instead of the generated summaries; if AUC does not fall, summarization is not the mechanism.

Watch

Extended reading notes

Core claim

xHAIM outperforms the HAIM baseline on all five tasks it evaluates. On the HAIM-MIMIC-MM dataset, average ROC AUC goes from 79.9% with HAIM to 90.3% with the best xHAIM variant, with per-task improvements of +13.5% for pleural effusion, +16.3% for cardiomegaly, +19.4% for pneumonia, +2.7% for 48-hour mortality, and +1.9% for length of stay. The paper attributes the gain to replacing noisy, averaged embeddings of full patient records with concise task-specific summaries produced by generative LLMs, and to fine-tuning ClinicalBERT on those summaries. It further claims that the same summaries, combined with the calibrated prediction and relevant medical knowledge, yield explanations that cite the exact documents supporting each prediction, and that an LLM-as-a-Judge evaluation of these explanations aligns with human annotation.

Load-bearing premise

The headline comparison assumes the HAIM baseline was evaluated under the same five-split protocol and that the entire improvement comes from the summarization step, not from the simultaneous switch to fine-tuned ClinicalBERT embeddings and Qwen2.5-VL image descriptions.

Editorial extensions

If this is right

  • If xHAIM's results hold, generative summarization becomes a drop-in front end for existing discriminative clinical predictors, not a replacement for them.
  • The largest AUC gains appear on narrative-dependent chest pathology tasks, implying that data curation matters most where clinical notes carry the signal.
  • Cited, document-grounded explanations can be produced automatically at scale, with LLM judge scores comparable to human annotations, shortening verification time for clinicians.
  • Because the pipeline can run on open-source models inside a hospital, the performance and explainability gains do not require sending patient data to external APIs.
  • Summaries that standardize free text may make the downstream model more robust to differences in note-writing style across institutions.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: the paper does not isolate the summarization step from the other changes in the pipeline, so the +10.4 average AUC should be read as the joint effect of fine-tuned ClinicalBERT, Qwen2.5-VL image descriptions, and curated summaries; an ablation replacing summaries with full notes would separate these.
  • Editorial inference: if the baseline HAIM numbers were carried over from the original publication rather than re-run on the same splits, the comparison is weaker than the paper presents; re-running HAIM under the xHAIM protocol would settle the magnitude of the gain.
  • Editorial inference: the explanation evaluation was calibrated on 50 examples per task and then scaled to 1,000; a natural next test is whether LLM-judge agreement with clinicians degrades on rare or atypical presentations.
  • Editorial inference: the same retrieval-summarize-predict-explain recipe could be carried to other clinical settings, such as outpatient notes or longitudinal EHR data, where the 'more data is better' assumption is even more costly.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript introduces xHAIM, an extension of the HAIM multimodal framework for ICU prediction. The pipeline (1) retrieves task-relevant text chunks using a hybrid BM25/SBERT score, (2) generates patient summaries with LLMs (Llama-3.3-70B, Qwen3-32B, and Qwen2.5-VL-72B for images), (3) fine-tunes ClinicalBERT on the summaries and integrates the resulting embeddings with tabular and time-series features, and (4) generates document-grounded explanations. On the HAIM-MIMIC-MM dataset, the authors report an average AUC improvement from 79.9% (HAIM baseline) to 90.3% (xHAIM-FT-Llama) across pleural effusion, cardiomegaly, pneumonia, 48-hour mortality, and length of stay. They also report manual and LLM-as-a-judge evaluations of explanation quality for two of the five tasks.

Significance. If established, the central claim that intelligent data curation via summarization yields large AUC gains while adding explainability would be an important contribution to medical AI. The paper uses a well-known public dataset, provides a concrete four-step pipeline, and attempts to validate explanations with both human annotators and an LLM judge. The strengths are the hybrid generative/discriminative framing, the use of open-source models, and the detailed description of the preprocessing steps. However, the evidence currently presented does not isolate the contribution of the curation step, and the apparent reuse of the original HAIM baseline undermines the headline comparison. The significance of the claimed 10.4-point gain is real if the comparison is valid, but the manuscript does not yet establish that validity.

major comments (4)
  1. [Results, Table 2] The HAIM Baseline row reports AUCs and standard errors (84.8±0.5, 81.1±0.2, 76.3±0.4, 82.0±0.2, 75.5±0.4) that match the original HAIM publication to the reported precision, and the manuscript never states that HAIM was re-run under the same five 80/20 splits used for xHAIM. Since split indices, per-stay scan selection, and preprocessing may differ, this baseline cannot support the claimed 79.9-to-90.3 improvement. The authors must re-run the original HAIM pipeline under the identical protocol and report the resulting AUCs and standard errors, or the headline comparison is not valid.
  2. [Methods, Multimodal Integration] xHAIM simultaneously changes three components relative to HAIM: (i) ClinicalBERT embeddings are fine-tuned on summaries instead of frozen on raw text, (ii) DenseNet121 image features are replaced by Qwen2.5-VL-72B text descriptions, and (iii) raw chunks are replaced by curated, summarized text. The observed gains cannot be attributed to 'intelligent data curation' without ablations that vary one component at a time. For example, report HAIM with fine-tuned ClinicalBERT on raw, non-summarized text, and report xHAIM without the chunk-retrieval step, to isolate the effect of summarization.
  3. [Methods, Finding Relevant Chunks] The anchor-sentence retrieval injects label-related vocabulary into the summaries: the anchors for pneumonia contain 'pneumonia,' 'consolidation,' and 'infiltrate,' and similar anchors are used for other outcomes. Because the selected chunks are then summarized and embedded by ClinicalBERT, the model's input features contain explicit mentions of the target condition, creating a circularity that can inflate AUC beyond genuine predictive signal. The authors should either use outcome-blind anchors (e.g., general clinical concepts) or demonstrate that classification performance does not rely on this lexical overlap, for instance by masking outcome terms in the retrieved chunks.
  4. [Results, Explainability Evaluation] The explanation-quality evaluation is limited to two conditions (pleural effusion and 48-hour mortality) with N=50 manual annotations per condition and N=1000 LLM-judge evaluations; Tables 3 and 4 do not cover cardiomegaly, pneumonia, or length of stay. The general claim that xHAIM provides clinically meaningful explanations across all tasks is therefore only partially supported. In addition, the text promises that supplementary information includes ablation studies, but no supplementary material is provided with this submission, and the code repository is not yet available, so the promised ablations and split definitions cannot be inspected.
minor comments (6)
  1. [General] The manuscript contains unresolved placeholder references to 'Appendix ??' in Sections 2 and 4; these must be filled in.
  2. [Results, Tables 3 and 4] No inter-annotator agreement metric is reported for the manual evaluations, so the claim of 'strong alignment with human judgment' by the LLM judge is not quantified.
  3. [Abstract / Results] The abstract mentions 'operative tasks,' but the evaluated tasks are chest pathology, mortality, and length of stay; the terminology should be clarified.
  4. [Methods, Finding Relevant Chunks] The hyperparameter alpha=0.5 in the hybrid score is fixed without sensitivity analysis; the authors should report whether results are robust to this choice.
  5. [Methods, Modalities Preprocessing] The description says one scan is randomly selected per ICU stay; the original HAIM scan-selection procedure should be stated explicitly so that comparability can be assessed.
  6. [Results, Table 1] Mortality and length of stay have identical cohort sizes (16,888); if this is due to a shared ground-truth definition, it should be stated in the text.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: xHAIM's AUC claim is compared against an external published baseline, and its components are not defined in terms of the target predictions.

full rationale

The paper's central claim is that xHAIM improves average AUC from 79.9% to 90.3% relative to the original HAIM framework on the same public HAIM-MIMIC-MM dataset. This is a benchmark comparison against an external, published baseline, not a derivation that reduces to its own inputs. The xHAIM models are trained and evaluated on held-out splits, and no fitted parameter is renamed as a prediction. The relevant-chunk retrieval step uses task-specific anchor sentences containing disease keywords such as 'pneumonia', which may inject label-related text into the summaries; this is a disclosed methodological choice and a potential confounding factor in attributing the gain specifically to summarization, but it does not make the prediction equivalent to the input by construction, because the model still must infer the presence or absence of the condition from clinical text. The explanation evaluation uses an LLM-as-a-judge calibrated against 50 human annotations per task, which provides an external anchor and is not purely self-referential. The HAIM baseline values appear to be taken from the original HAIM paper rather than re-run on identical splits, which is a serious experimental-validity concern but not a circularity, since the baseline is an independent published result. Overall, the derivation chain is self-contained and no load-bearing step reduces to a self-citation or to the target variable by definition.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The framework rests on assumptions that LLM summaries preserve clinically relevant information without material hallucination, that task keyword-based retrieval selects the relevant content, and that the MIMIC-derived labels are correct. These are domain assumptions, not derived from first principles. There are no new physical or ontological entities.

free parameters (4)
  • alpha (hybrid score weight) = 0.5 (hand-chosen); no sensitivity analysis is reported
    Weights BM25 and SBERT similarity in the chunk retrieval score. This choice is not tuned and no robustness check is provided.
  • top-k chunks retained per modality = Not specified in the manuscript.
    The number of chunks kept for summarization is not stated, yet it controls how much information reaches the LLM summary and thus the prediction.
  • Task-specific anchor phrases = Hand-crafted per task (e.g., 'pneumonia', 'consolidation', 'infiltrate' for pneumonia).
    These phrases define what counts as task-relevant for the retrieval step and inject target-related vocabulary into the summaries.
  • Summarization and explanation prompts = Not provided in the manuscript.
    The exact LLM prompts for summary and explanation generation are not disclosed, which are free design choices that affect outputs.
assumptions (4)
  • domain assumption LLM-generated summaries preserve all task-relevant clinical information and do not materially distort or hallucinate facts.
    Invoked in Step 2 (Generating Task-Specific Summaries); if summaries drop or alter key findings, prediction and explanations degrade.
  • domain assumption Semantic similarity between chunks and hand-chosen anchor phrases identifies the clinically relevant portions of the record.
    Invoked in Step 1 (Finding Relevant Chunks); the retrieval defines the information budget for prediction.
  • domain assumption The HAIM-MIMIC-MM ground-truth labels for pleural effusion, cardiomegaly, pneumonia, mortality, and LOS are accurate and consistently defined.
    The dataset is used as the evaluation benchmark; label errors would bias all reported AUCs.
  • domain assumption Discriminative models trained on summary embeddings generalize from the 80/20 split to the broader ICU population.
    Standard supervised-learning assumption; external validation is acknowledged as future work.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Holistic Artificial Intelligence in Medicine; improved performance and explainability." pith.science (2026). https://pith.science/paper/RAHIISBW

@misc{pith2026250700205,
  author       = {Pith},
  title        = {Pith review of: Holistic Artificial Intelligence in Medicine; improved performance and explainability},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RAHIISBW}},
  note         = {Machine review of arXiv:2507.00205}
}
read the original abstract

With the increasing interest in deploying Artificial Intelligence in medicine, we previously introduced HAIM (Holistic AI in Medicine), a framework that fuses multimodal data to solve downstream clinical tasks. However, HAIM uses data in a task-agnostic manner and lacks explainability. To address these limitations, we introduce xHAIM (Explainable HAIM), a novel framework leveraging Generative AI to enhance both prediction and explainability through four structured steps: (1) automatically identifying task-relevant patient data across modalities, (2) generating comprehensive patient summaries, (3) using these summaries for improved predictive modeling, and (4) providing clinical explanations by linking predictions to patient-specific medical knowledge. Evaluated on the HAIM-MIMIC-MM dataset, xHAIM improves average AUC from 79.9% to 90.3% across chest pathology and operative tasks. Importantly, xHAIM transforms AI from a black-box predictor into an explainable decision support system, enabling clinicians to interactively trace predictions back to relevant patient data, bridging AI advancements with clinical utility.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

45 extracted references · 20 canonical work pages

  1. [1]

    BMC Medical Informatics and Decision Making 20(1), 310 (2020) https://doi.org/10.1186/s12911-020-01332-6

    Amann, J., Blasimme, A., Vayena, E., Frey, D., Madai, V.I., the Precise4Q con- sortium: Explainability for artificial intelligence in healthcare: a multidisciplinary perspective. BMC Medical Informatics and Decision Making 20(1), 310 (2020) https://doi.org/10.1186/s12911-020-01332-6 . Accessed 2025-06-17

    doi:10.1186/s12911-020-01332-6 2020

  2. [2]

    Information Fusion 79, 263–278 (2022) https://doi

    Holzinger, A., Dehmer, M., Emmert-Streib, F., Cucchiara, R., Augenstein, I., Ser, J.D., Samek, W., Jurisica, I., D ´ ıaz-Rodr ´ ıguez, N.: Information fusion as an integrative cross-cutting enabler to achieve robust, explainable, and trustworthy medical artificial intelligence. Information Fusion 79, 263–278 (2022) https://doi. org/10.1016/j.inffus.2021.1...

    doi:10.1016/j.inffus.2021.10.007 2022

  3. [3]

    ACM computing surveys (CSUR) 54(6), 1–35 (2021)

    Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., Galstyan, A.: A survey on bias and fairness in machine learning. ACM computing surveys (CSUR) 54(6), 1–35 (2021). Publisher: ACM New York, NY, USA

    work page 2021

  4. [4]

    Nature 620(7972), 172–180 (2023) https://doi.org/10.1038/ s41586-023-06291-2

    Singhal, K., Azizi, S., Tu, T., Mahdavi, S.S., Wei, J., Chung, H.W., Scales, N., Tanwani, A., Cole-Lewis, H., Pfohl, S., Payne, P., Seneviratne, M., Gam- ble, P., Kelly, C., Babiker, A., Sch¨ arli, N., Chowdhery, A., Mansfield, P., Demner-Fushman, D., Arcas, B., Webster, D., Corrado, G.S., Matias, Y., Chou, K., Gottweis, J., Tomasev, N., Liu, Y., Rajkomar...

    work page 2023

  5. [5]

    Nature Medicine 31(3), 943–950 (2025) https://doi.org/10.1038/s41591-024-03423-7

    Singhal, K., Tu, T., Gottweis, J., Sayres, R., Wulczyn, E., Amin, M., Hou, L., Clark, K., Pfohl, S.R., Cole-Lewis, H., Neal, D., Rashid, Q.M., Schaekermann, M., Wang, A., Dash, D., Chen, J.H., Shah, N.H., Lachgar, S., Mansfield, P.A., Prakash, S., Green, B., Dominowska, E., Arcas, B., Tomaˇ sev, N., Liu, Y., Wong, R., Semturs, C., Mahdavi, S.S., Barral, J...

    2025

  6. [6]

    Saab, K., Tu, T., Weng, W.-H., Tanno, R., Stutz, D., Wulczyn, E., Zhang, F., Strother, T., Park, C., Vedadi, E., Chaves, J.Z., Hu, S.-Y., Schaekermann, M., Kamath, A., Cheng, Y., Barrett, D.G.T., Cheung, C., Mustafa, B., Palepu, A., McDuff, D., Hou, L., Golany, T., Liu, L., Alayrac, J.-b., Houlsby, N., Tomasev, N., Freyberg, J., Lau, C., Kemp, J., Lai, J....

    arXiv doi:10.48550/arxiv.2404.18416 2024

  7. [7]

    Nature Medicine 29(8), 1930– 1940 (2023) https://doi.org/10.1038/s41591-023-02448-8

    Thirunavukarasu, A.J., Ting, D.S.J., Elangovan, K., Gutierrez, L., Tan, T.F., Ting, D.S.W.: Large language models in medicine. Nature Medicine 29(8), 1930– 1940 (2023) https://doi.org/10.1038/s41591-023-02448-8 . Publisher: Nature Publishing Group. Accessed 2025-06-17

    doi:10.1038/s41591-023-02448-8 2023

  8. [8]

    Wang, Z., Wang, H., Danek, B., Li, Y., Mack, C., Poon, H., Wang, Y., Rajpurkar, P., Sun, J.: A Perspective for Adapting Generalist AI to Specialized Medical AI Applications and Their Challenges. arXiv. arXiv:2411.00024 (2024). http://arxiv. org/abs/2411.00024 Accessed 2024-11-10

    work page Pith review arXiv 2024

Show all 45 references
  1. [9]

    ArXiv (2023)

    Bubeck, S., Chadrasekaran, V., Eldan, R., Gehrke, J., Horvitz, E., Kamar, E., Lee, P., Lee, Y.T., Li, Y., Lundberg, S., et al.: Sparks of artificial general intelligence: Early experiments with gpt-4. ArXiv (2023)

    2023

  2. [10]

    Advances in neural information processing systems 33, 1877–1901 (2020)

    Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J.D., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al.: Language models are few-shot learners. Advances in neural information processing systems 33, 1877–1901 (2020)

    2020

  3. [11]

    Journal of the American Medical Informatics Association, 065 (2025)

    Poon, E.G., Lemak, C.H., Rojas, J.C., Guptill, J., Classen, D.: Adoption of arti- ficial intelligence in healthcare: survey of health system priorities, successes, and challenges. Journal of the American Medical Informatics Association, 065 (2025). Publisher: Oxford University Press

    2025

  4. [12]

    Research Report, American Medical Association (2024)

    American Medical Association: Augmented Intelligence in Medicine: 2024 Physi- cian Survey Results. Research Report, American Medical Association (2024)

    2024

  5. [13]

    JAMA 333(4), 319–328 (2025) https://doi.org/10.1001/jama.2024.21700

    Bedi, S., Liu, Y., Orr-Ewing, L., Dash, D., Koyejo, S., Callahan, A., Fries, J.A., Wornow, M., Swaminathan, A., Lehmann, L.S., Hong, H.J., Kashyap, M., Chaurasia, A.R., Shah, N.R., Singh, K., Tazbaz, T., Milstein, A., Pfef- fer, M.A., Shah, N.H.: Testing and Evaluation of Heal...

    2025

  6. [14]

    BMC medicine 17, 1–9 (2019)

    Kelly, C.J., Karthikesalingam, A., Suleyman, M., Corrado, G., King, D.: Key challenges for delivering clinical impact with artificial intelligence. BMC medicine 17, 1–9 (2019)

    2019

  7. [15]

    BMC Medical Ethics 25(1), 104 (2024)

    Freyer, N., Groß, D., Lipprandt, M.: The ethical requirement of explainability 14 for AI-DSS in healthcare: a systematic review of reasons. BMC Medical Ethics 25(1), 104 (2024). Publisher: Springer

    2024

  8. [16]

    Publisher: Springer

    Abgrall, G., Holder, A.L., Chelly Dagdia, Z., Zeitouni, K., Monnet, X.: Should AI models be explainable to clinicians? Critical Care 28(1), 301 (2024). Publisher: Springer

    2024

  9. [17]

    PhysioNet (2022)

    Soenksen, L.R., Ma, Y., Zeng, C., Boussioux, L.D.J., Villalobos Carballo, K., Na, L., Wiberg, H., Li, M., Fuentes, I., Bertsimas, D.: Code for generating the HAIM multimodal dataset of MIMIC-IV clinical data and x-rays (version 1.0.1). PhysioNet (2022). https://doi.org/10.1302...

    2022 doi

  10. [18]

    Scientific data 3(1), 1–9 (2016)

    Johnson, A.E., Pollard, T.J., Shen, L., Lehman, L.-w.H., Feng, M., Ghassemi, M., Moody, B., Szolovits, P., Anthony Celi, L., Mark, R.G.: MIMIC-III, a freely acces- sible critical care database. Scientific data 3(1), 1–9 (2016). Publisher: Nature Publishing Group

    2016

  11. [19]

    Scientific data 10(1), 1 (2023)

    Johnson, A.E., Bulgarelli, L., Shen, L., Gayles, A., Shammout, A., Horng, S., Pollard, T.J., Hao, S., Moody, B., Gow, B., et al.: MIMIC-IV, a freely accessible electronic health record dataset. Scientific data 10(1), 1 (2023). Publisher: Nature Publishing Group UK London

    2023

  12. [20]

    In: Artificial Intelligence and Statistics, pp

    McMahan, B., Moore, E., Ramage, D., Hampson, S., Arcas, B.A.: Communication-efficient learning of deep networks from decentralized data. In: Artificial Intelligence and Statistics, pp. 1273–1282 (2017). PMLR

    2017

  13. [21]

    Cell Reports Medicine 5(2) (2024)

    Teo, Z.L., Jin, L., Liu, N., Li, S., Miao, D., Zhang, X., Ng, W.Y., Tan, T.F., Lee, D.M., Chua, K.J., et al.: Federated machine learning in healthcare: A systematic review on clinical applications and technical architecture. Cell Reports Medicine 5(2) (2024)

    2024

  14. [22]

    NPJ digital medicine 3(1), 119 (2020)

    Rieke, N., Hancox, J., Li, W., Milletari, F., Roth, H.R., Albarqouni, S., Bakas, S., Galtier, M.N., Landman, B.A., Maier-Hein, K., et al.: The future of digital health with federated learning. NPJ digital medicine 3(1), 119 (2020)

    2020

  15. [23]

    https://arxiv.org/abs/1904.05342

    Huang, K., Altosaar, J., Ranganath, R.: ClinicalBERT: Modeling Clinical Notes and Predicting Hospital Readmission (2020). https://arxiv.org/abs/1904.05342

    2020 arXiv

  16. [24]

    Zhang, X., Acosta, J.N., Zhou, H.-Y., Rajpurkar, P.: Uncovering Knowledge Gaps in Radiology Report Generation Models through Knowledge Graphs. arXiv. arXiv:2408.14397 (2024). http://arxiv.org/abs/2408.14397 Accessed 2024-11-10

    2024 arXiv

  17. [25]

    npj Digital Medicine 5(1), 149 (2022) https://doi.org/10.1038/s41746-022-00689-4

    Soenksen, L.R., Ma, Y., Zeng, C., Boussioux, L., Villalobos Carballo, K., Na, L., Wiberg, H.M., Li, M.L., Fuentes, I., Bertsimas, D.: Integrated multimodal artificial intelligence framework for healthcare applications. npj Digital Medicine 5(1), 149 (2022) https://doi.org/10.1...

    2022 doi

  18. [26]

    Research square, 3 (2024)

    Xie, Q., Chen, Q., Chen, A., Peng, C., Hu, Y., Lin, F., Peng, X., Huang, J., Zhang, J., Keloth, V., et al.: Me-llama: Foundation large language models for medical applications. Research square, 3 (2024)

    2024

  19. [27]

    NPJ Digital Medicine 7(1), 20 (2024)

    Savage, T., Nayak, A., Gallo, R., Rangan, E., Chen, J.H.: Diagnostic reasoning prompts reveal the potential for large language model interpretability in medicine. NPJ Digital Medicine 7(1), 20 (2024)

    2024

  20. [28]

    Chen, C., Yu, J., Chen, S., Liu, C., Wan, Z., Bitterman, D., Wang, F., Shu, K.: ClinicalBench: Can LLMs Beat Traditional ML Models in Clinical Prediction? arXiv preprint arXiv:2411.06469 (2024)

    2024 arXiv

  21. [29]

    medRxiv (2024)

    Brown, K.E., Yan, C., Li, Z., Zhang, X., Collins, B.X., Chen, Y., Clayton, E.W., Kantarcioglu, M., Vorobeychik, Y., Malin, B.A.: Not the models you are look- ing for: Traditional ML outperforms LLMs in clinical prediction tasks. medRxiv (2024)

    2024

  22. [30]

    arXiv preprint arXiv:2503.05777 (2025)

    Kim, Y., Jeong, H., Chen, S., Li, S.S., Lu, M., Alhamoud, K., Mun, J., Grau, C., Jung, M., Gameiro, R., et al.: Medical hallucinations in foundation models and their impact on healthcare. arXiv preprint arXiv:2503.05777 (2025)

    2025

  23. [31]

    Liu, N.F., Lin, K., Hewitt, J., Paranjape, A., Bevilacqua, M., Petroni, F., Liang, P.: Lost in the Middle: How Language Models Use Long Contexts. arXiv. arXiv:2307.03172 [cs] (2023). https://doi.org/10.48550/arXiv.2307.03172 . http: //arxiv.org/abs/2307.03172 Accessed 2025-02-26

    2023 doi arXiv

  24. [32]

    arXiv preprint arXiv:2305.14251 (2023)

    Min, S., Krishna, K., Lyu, X., Lewis, M., Yih, W.-t., Koh, P.W., Iyyer, M., Zettle- moyer, L., Hajishirzi, H.: Factscore: Fine-grained atomic evaluation of factual precision in long form text generation. arXiv preprint arXiv:2305.14251 (2023)

    2023 arXiv

  25. [33]

    In: Ku, L.-W., Martins, A., Srikumar, V

    Xiong, G., Jin, Q., Lu, Z., Zhang, A.: Benchmarking Retrieval-Augmented Generation for Medicine. In: Ku, L.-W., Martins, A., Srikumar, V. (eds.) Findings of the Association for Computational Linguistics ACL 2024, pp. 6233–6251. Association for Computational Linguistics, Bangko...

    2024 doi

  26. [34]

    npj Digital Medicine 8(1), 175 (2025)

    Takita, H., Kabata, D., Walston, S.L., Tatekawa, H., Saito, K., Tsujimoto, Y., Miki, Y., Ueda, D.: A systematic review and meta-analysis of diagnostic perfor- mance comparison between generative AI and physicians. npj Digital Medicine 8(1), 175 (2025). Publisher: Nature Publis...

    2025

  27. [35]

    Wornow, M., Thapa, R., Steinberg, E., Fries, J.A., Shah, N.H.: EHRSHOT: An EHR Benchmark for Few-Shot Evaluation of Foundation Models. arXiv. arXiv:2307.02028 [cs] (2023). https://doi.org/10.48550/arXiv.2307.02028 . http: 16 //arxiv.org/abs/2307.02028 Accessed 2025-03-24

    2023 doi arXiv

  28. [36]

    Fleming, S.L., Lozano, A., Haberkorn, W.J., Jindal, J.A., Reis, E.P., Thapa, R., Blankemeier, L., Genkins, J.Z., Steinberg, E., Nayak, A., Patel, B.S., Chi- ang, C.-C., Callahan, A., Huo, Z., Gatidis, S., Adams, S.J., Fayanju, O., Shah, S.J., Savage, T., Goh, E., Chaudhari, A....

    2023 doi arXiv

  29. [37]

    ACM transactions on intelligent systems and technology 15(3), 1–45 (2024)

    Chang, Y., Wang, X., Wang, J., Wu, Y., Yang, L., Zhu, K., Chen, H., Yi, X., Wang, C., Wang, Y.,et al.: A survey on evaluation of large language models. ACM transactions on intelligent systems and technology 15(3), 1–45 (2024). Publisher: ACM New York, NY

    2024

  30. [38]

    arXiv preprint arXiv:2302.04023 (2023)

    Bang, Y., Cahyawijaya, S., Lee, N., Dai, W., Su, D., Wilie, B., Lovenia, H., Ji, Z., Yu, T., Chung, W., et al.: A multitask, multilingual, multimodal eval- uation of chatgpt on reasoning, hallucination, and interactivity. arXiv preprint arXiv:2302.04023 (2023)

    2023 arXiv

  31. [39]

    Miura, Y., Zhang, Y., Tsai, E.B., Langlotz, C.P., Jurafsky, D.: Improving Factual Completeness and Consistency of Image-to-Text Radiology Report Generation. arXiv. arXiv:2010.10042 (2021). http://arxiv.org/abs/2010.10042 Accessed 2024- 11-10

    2021 arXiv

  32. [40]

    In: Proceedings of the 61st Annual Meeting of the Associ- ation for Computational Linguistics (Volume 2: Short Papers), pp

    Chen, Z., Varma, M., Wan, X., Langlotz, C., Delbrouck, J.-B.: Toward expanding the scope of radiology report summarization to multiple anatomies and modalities. In: Proceedings of the 61st Annual Meeting of the Associ- ation for Computational Linguistics (Volume 2: Short Paper...

    2023 arXiv

  33. [41]

    5-vl technical report

    Bai, S., Chen, K., Liu, X., Wang, J., Ge, W., Song, S., Dang, K., Wang, P., Wang, S., Tang, J., et al.: Qwen2. 5-vl technical report. arXiv preprint arXiv:2502.13923 (2025)

    2025 arXiv

  34. [42]

    Foundations and Trends ® in Information Retrieval 3(4), 333–389 (2009)

    Robertson, S., Zaragoza, H., et al.: The probabilistic relevance framework: Bm25 and beyond. Foundations and Trends ® in Information Retrieval 3(4), 333–389 (2009)

    2009

  35. [43]

    arXiv preprint arXiv:1908.10084 (2019)

    Reimers, N., Gurevych, I.: Sentence-bert: Sentence embeddings using siamese bert-networks. arXiv preprint arXiv:1908.10084 (2019)

    2019 arXiv

  36. [44]

    Grattafiori, A., Dubey, A., Jauhri, A., Pandey, A., Kadian, A., Al-Dahle, A., Let- man, A., Mathur, A., Schelten, A., Vaughan, A., Yang, A., Fan, A., Goyal, A., Hartshorn, A., Yang, A., Mitra, A., Sravankumar, A., Korenev, A., Hinsvark, 17 A., Rao, A., Zhang, A., Rodriguez, A....

    2024 doi arXiv

  37. [45]

    Yang, A., Li, A., Yang, B., Zhang, B., Hui, B., Zheng, B., Yu, B., Gao, C., Huang, C., Lv, C., Zheng, C., Liu, D., Zhou, F., Huang, F., Hu, F., Ge, H., Wei, H., Lin, H., Tang, J., Yang, J., Tu, J., Zhang, J., Yang, J., Yang, J., Zhou, J., Zhou, J., Lin, J., Dang, K., Bao, K., ...

    2025 doi arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.